“Persons pretending to forecast the future shall be considered disorderly” – New York Crime Code
In today’s changing markets, forecasting is becoming ever more important. “Time to expand”, “steady as she goes”, or “time to batten down the hatches”, are just a few of the key decisions that are fed by the forecasting engine of a company.
Are you on the “forecasting is an art and not a science” side of the debate? I believe forecasting is a combination of reliable, process driven systems, tools to summarize the data, and most importantly, the proper interpretation of the data. This is a combination of art and science and relies on both the experience of the forecaster and the ability to see trends in numbers. Good forecasters will always listen very closely to their ‘gut feel’ as this often is a signal to dig deeper into certain aspects of the forecast.
It is also important to catch patterns in forecasting within a company. Examples are ‘salespeople always forecast too high’, ‘big orders always come in later than we expect’, and even ‘we can never seem to forecast the last few weeks of the quarter’. All of these sound quite like systematic problems in the methodology and can normally be solved by a detailed review of the system, an understanding of the compensation model of the staff who are forecasting, management ‘pressures’ on the sales organisation and a whole host of other more specific areas than can cause a ‘consistent error’ in forecasting accuracy.
Ever had to forecast in Canadian Dollars when customers buy in local currency (e.g. US Dollars)? I remember in the past having the challenge of trying to grow revenues faster than the negative changes in the exchange rate. With today’s US/Canadian exchange rate hovering around parity it is wise to get the CEO, CFO and VP of Sales aligned as to forecasted exchange rates over time.
Of course forecasting will always have a degree of variability, so management needs to be clear on what level of accuracy is actually required by the Company to run successfully. A closed-loop system should be put in place to measure and monitor performance, and corrective actions applied when accuracy falls outside a pre-determined range.
So, go on, set a forecast accuracy target and start measuring yourself against that accuracy – it could be quite illuminating!